The discrepant results noted by the La Trobe/Monash, Sussex, Open, and Berkeley University memory research groups employing the passive avoidance task (PAT) with the chick indicate that some of these differences may be due to differences in training procedures between the various groups. The procedures employed by each group were replicated as closely as possible and compared using the same strain of chick. Higher levels of pecking and lower training latencies were observed in the La Trobe/Monash chicks. Improved training latency was observed following a change in the day of experimentation with Sussex chicks, and in white light in Open chicks. A powerful reinforcing effect was observed when pretraining was conducted with a wet lure across multiple pretraining trials, indicating the importance of this difference between groups. The observed impact on chick performance of these variables indicates that differences between PAT procedures may have important interactive effects with aversive learning. The results of data gathered using different procedures may only be superficially comparable, and further investigation of the effects of these variables is indicated.