In August 1998, 3000 Atlantic salmon Salmo salar L. parr were divided into 7 groups with 2 replicates. Every 6 wk until March of the following year 1 group was vaccinated. One group was held as an unvaccinated control. The fish were transferred to seawater in May 1999, and slaughtered in February 2000. Temperature, fish size and photoperiod at vaccination, and the time between vaccination and sea transfer thus varied among the groups. In all vaccinated groups, growth was reduced for 1 to 2 mo following vaccination. Intra-abdominal lesions developed faster, and stabilised at a higher level in the groups vaccinated early at the highest temperature and the smallest fish size. Growth in seawater was influenced by the time of vaccination. At the end of the experiment, the group vaccinated last (MAR) was the heaviest of the vaccinated groups (4.0 kg), and the group vaccinated first, i.e. in August (AUG) was smallest (3.2 kg). Growth rate in seawater differed only in the summer when specific growth rate was above 1.45 in all groups. There was a correlation between adhesion, condition factor and number of weeks from vaccination to sea transfer. The AUG group had the highest condition factor, with a top level of 1.64 in autumn, and this group also displayed the highest incidence of deformed vertebra. The experiment shows that side effects of vaccination can be significantly reduced when planning the vaccination strategy, by taking environmental factors and fish biology into consideration.