The monthly prison population for each institution was the population at the end of the month, beginning with January 1980 and ending with August 1989. The disciplinary infractions at the institutions were divided into those that involved violence or the threat of violence and those that did not. Across all institutions, findings show no consistent pattern of effects for density. Even within institutions, there was no consistent pattern across time. Multivariate analyses that held other factors constant indicate that institutional density and rate of change of prison populations had statistically significant effects on both types of infractions, although the size of the effects was very small. Additional analyses indicate that these effects were particularly true for prisons for youth. Overall, the results indicate that institutional factors other than density were the strongest predictors of infractions. 2 tables and 23 references