Given its importance as an example of psychic functioning, I thought it was worth looking as this event in some detail, using documents from the declassified archive.
The earliest mention of this experiment I can find is an appendix to a meeting held on 31 May 1974, where basic protocols were laid down: that the sessions would be “guided” by someone knowledgable about the site; but giving as little cueing as possible and then, over a period of a few days, introduce more information about the site to Pat Price “recognizing that the additional information supplied to the viewer at this point is at the cost of calibration within the experiment.” 
The initial data given to him at the start of the experiment would be a map with the site marked on it and some drawings to help Price identify the location when he remote views it.
On the 17 June 1974, the NPIC (the National Photographic Intelligence Center, who’d be supplying the data against which Price’s statements would be compared) and the NED (can’t find this acronym!) were both contacted for help in evaluating the data. It seems that the source was to be disguised before judging, and instead of overtly being from a psychic, it was going to be attributed to a janitor: someone with little technical knowledge but open access to URDF-3.
Semipalatinsk URDF-3, from top of gantry crane.
Also at this time, a member of ORD (Office of Research & Development) visited SRI for two days, 17-18 June. During this time it was alluded that the SRI program was in danger of being closed, unless certain targets were met.
This was followed by another visit from the ORD on the 28 June, and the report on this visit stated that “the progress was disappointing” and “it appears that a pot pourri of small experiments were to have been conducted on an ad hoc basis on whatever subjects were around.”
I mention these two visits to put the URDF-3 session into some kind of context: the SRI project was falling behind schedule and Targ and Puthoff must've been keenly aware of the pressure from their sponsors.
On 27 and 28 June two meetings were held, the second one by telephone, to fine tune the experimental methods. The conclusions to this meeting were written up in notes on the appendix from the meeting held in May. 
Now the protocol specified that the person who knew about the target would not be physically present during the sessions, nor would Price be given any drawings before the sessions began.
More details regarding the protocol were given in a memo dated 3 July 1974 (ie, five days before the sessions were to begin). This memo stated that there would be three stages to the protocol. It also mentioned that, during stage two, Price would be told that the sponsor for this experiment was a government agency.
The experiment itself ran from 9-12 July 1974. The majority of information about the details of this period of time comes from a report dated 4 December 1975 “An Analysis Of A Remote-Viewing Experiment of URDF-3” by D. Stillman, a nuclear analyst from the Los Alamos Scientific Laboratory, working from sketches and tape recordings.  It’s worth noting that the plan to disguise the psychic nature of the data had been dropped but Stillman says in his introduction that, if anything, he would like psychic functioning to be true.
At the start of the experiment, Price was given the map co-ordinates (50°09'59"N 78°22'22"E), was shown the site’s position on a number of maps, and told that it was 60 miles WSW of Semipalatinsk and 25 to 30 SW of a river and he was told it is a scientific military research and test area. Once he began remote viewing, he found the river and then the location. Stillman reports:
“He says the area he’s looking at has low one-story buildings that are partially dug into the ground giving the effect (as seen at ground level) of very short, squatty buildings, whereas they are actually fairly roomy on this inside. This description could very well describe a first look at the Operations Area at URDF-3.”
From Google Maps, imagery dated 2005
Price then started describing tests being run that were related to the space program, followed by drawing a map of the research facility and the area which was incorrect, apart from the information already given to him.
Most of the data from Price in day one is wrong, and he shows an inclination to be led by the questioner, such as when the scientist in the room with Price during the session (probably Russell Targ) makes a suggestion as to what something he’s describing might be, Price says “You could be right,” before the scientist quickly reminds Price that he knows nothing about the target.
The famous crane hit did not happen during day one, but in the evening. Midway through the first afternoon session, Price mentioned a crane and some low boy trucks. Later that day, the person knowledgeable about URDF-3 (possibly Ken Kress) must have spoken to SRI because Hal Puthoff called Pat Price with instructions to draw crane and the security fence around the perimeter, which Pat Price had also mentioned in passing on day one.
The drawing of the crane was handed in on the evening of day one, and the tapes from day two (10 July) begin with a discussion of them, mentioning the size of the crane: 150 feet in height. Stillman wrote: “He [Price] said he didn’t realize how large the gantry crane was until he saw a man walking by one of the crane wheels.” And Stillman followed this with the observation “It seems inconceivable to imagine how he could draw such a likeness to the actual crane at URDF-3 unless: 1) he actually saw it through remote viewing, or 2) he was informed of what to draw by someone knowledgeable of URDF-3”
However, Stillman did not seem to have noticed that, until then, Price had frequently described features of a considerable size, such as an array of telegraph poles, a village, an airstrip, etc. It looks as if Price was working from the assumption that his work would be compared to aerial photography and the only data available would be something visible from a distance. As such, when he was given a clue that he had described something US intelligence knew to exist, he changed its description to one of great size.
Price’s sketch of the perimeter fence is only of a small section, and he doesn’t mention that there are actually four fences at URDF-3. However, Stillman drew in a perimeter fence on one of Price’s sketches because he could see that the shape was broadly similar to the fences around URDF-3.
Despite the success of the crane sketch, the rest of Price’s work on day two is largely unremarkable. Since his description of the huge crane did not match his earlier description of the building it interacts with, he spent a lot of time devising a complicated system including a second, smaller gantry crane, as well as drawing a large domed building and cement silo that Stillman says don’t exist at URDF-3.
Apart from the crane, there is another hit often attributed to the session: a large steel sphere under construction. It is not in the report by Stillman, which only covers the first two days in any detail. The third and fourth days supplied a lot of data but nothing that could be checked at that time.
So it must be from these two days, 11 and 12 July, when Pat Price described and drew large steel spheres mid-manufacture, in the form of individual gores, to be welded together. It is claimed that “Pat Price had been right, and he had described the spheres and the special welding techniques before anyone in the United States knew they existed.” 
The existence of these spheres was made public in a dramatic fashion in May 1977. Major General George J. Keegan, head of Air Force Intelligence had been trying (unsuccessfully) to convince the US military that the Soviets were building a particle beam weapon at Semipalatinsk for some years. When he retired, he went public with his suspicions and spoke to a writer from Aviation Weekly.  This sparked a public debate on the issue, with President Carter, the Defense Secretary, the CIA and the DIA all speaking out against the retired Major General’s conclusions.
While other research (and, eventually, complete disclosure from the Soviets) discredited the particle beam weapon theory, what was important to SRI was the detail about the steel spheres, constructed piece by piece, using a special welding method. Just as Pat Price had described.
It’s a shame that Stillman did not mention this or include any of day three’s six sketches in his report. Regrettably, the other analysis of Price’s work (W.T.Strand, “Memorandum for the Director, Office of Technical Service; Subject: Evaluation of Data on Semipalatinsk Unidentified R&D Facility No. 3, USSR; 20 August 1974”) is unavailable.
This means the earliest mention we have of this success is in Winter 1977, in a paper called “Parapsychology in Intelligence: A Personal Review and Conclusions” by Ken Kress, the CIA agent managing this particular experiment. He only mentions it in passing, though.
“In general, most of Price’s data were wrong or could not be evaluated. He did, nevertheless, produce some amazing description, like buildings then under construction, spherical tank sections, and the crane.” 
Price’s sketch of these sphere was published in 1983 in the heavily censored report Grill Flame Operational Tasks. On page seven, there is a sketch of metallic strips in the shape of a sphere segment, labelled “Sphere Fabrication”. 
There are a few issues with this as a piece of evidence. The first and most obvious is the date on the sketch. The handwritten date reads “6-18-74” which, if accurate, would mean it was drawn about one month before the URDF-3 work began. The other date, the rubber-stamped one, reads “7-18-74” which places it six days after the URDF-3 sessions were completed. Either way, this drawing does not appear to come from the four-day experiment. Additionally, throughout the sketches reproduced in Stillman’s report, whenever Price makes lengthy notes about a feature he does so in cursive writing, but in this sketch he has switched to writing in upper case.
There are the words “from prior” next to the handwritten date, so it’s possible that this is a redrawing of a sketch that SRI no longer had. This is likely since the session notes and tapes were not kept at SRI but sent off to be analysed. There is a letter from SRI to the CIA in 1986 asking for these to be returned but the reply stated that, apart from Stillman’s report, they couldn’t locate the other material. If this sketch is a reconstruction, one has to wonder how accurate it is and what was it based on.
The final problem with this sketch is related to the one immediately above it. Two sketches are labelled “Cyclinder clusters,” I assume to resemble the cluster of cylindrical tanks seen in photos of Semipalatinsk.
However, in Stillman’s report, this sketch is described as a “cement silo-like building” and not related to a cluster of cylinders at all. The details of Price’s descriptions have been removed in order to emphasise the similarity of the overall shape. If this sketch has been mis-labelled to increase the apparent level of success, has the sketch of the sphere sections also undergone a similar process?
Pat Price’s remote viewing of URDF-3 is frequently cited as a great success of psychic functioning. The dramatic similarity of Price’s drawing of a gantry crane with the sketch based on photographic intel makes it a popular choice when listing evidence for ESP. But this needs to be put into some kind of context: Price generated a great deal of data. Stillman was given about four hours of taped conversation from days one and two, seventy-nine pages of transcripts from days three and four, and a total of thirty sketches.
Furthermore, Price was not completely blind to the nature of the target and also someone knowledgeable about the location specifically asked him to draw the crane. Given this, it is hardly surprising that, on occasion, Price described features that could reasonably be compared to the actual location.
The final word on the issue should go to Stillman who, in his conclusion, wrote:
“In trying to determine the validity of this remote viewing experiment, the worth of the data to the eventual user has to be considered. If the user had no way of checking, how could he differentiate the fact from the fiction? In the case of URDF-3, the only positive evidence of the rail-mounted gantry was far outweighed by the large amount of negative evidence noted in the body of this analysis.”
 Appendix I: Suggested Protocol for Operational Remote-Viewing Exercise dated 31 May 1974 with handwritten notes added on or after 28 June
 D. Stillman (1975) An Analysis Of A Remote-Viewing Experiment of URDF-3
 Schnabel, J. (2011)Remote Viewers: The Secret History of America’s Psychic Spies, Random House Publushing Group
 Detailed summary of Aviation Weekly article.
 Kress, K.A. (1977) Parapsychology in Intelligence: A personal review and conclusions”
 Puthoff, H.E., Lavelle, L.A. (1983) "Project Grill Flame Operational Tasks"