The provision of a socialisation and training programme for dogs can lead to an improvement in the quality of the data that are produced from scientific procedures, and also to an improvement in animal welfare. A reduction in the number of animals needed to produce statistically significant data can result from decreasing the variability of the experimental data. The effects of a dog's behaviour can be a major source of random variability. A well socialised, habituated and trained dog should be calm during experimental procedures and, for example, during an ECG measurement, thus providing good quality data. A fearful, poorly socialised dog may also appear to be well-behaved on an examination table, and during an ECG measurement, because of the freezing response. However, there is likely to be a difference in the level of stress that these two individuals experience during the procedures. The stress response can have an impact on vital physiological parameters, such as heart rate. The variability in these parameters and the behaviour exhibited within a group of socialised, habituated and trained dogs that have been well prepared for experimental procedures, should be less than the variability present within a similar group of dogs that have not been prepared for these procedures. This paper describes two socialisation programmes, which were designed in order to compare the heart rates and behaviour of dogs which had received different degrees of socialisation, habituation and training. The behaviour of small groups of dogs from this study was compared with that of dogs on a standard socialisation programme, by using a simple, reproducible behavioural score scheme. The heart rate of the dogs was also measured. The results showed that there was little difference in heart rate between the groups, but that there were significant differences in the scores for key behaviours. There was evidence of a decrease in the variability of the behavioural scores for the groups of dogs that had undergone an intensive socialisation programme. Therefore, a socialisation programme can have a significant effect on behaviour and welfare, and has the potential to improve the quality of the data that are recorded.