The purposes of this study were to develop a locus of control scale specific to vocational needs and to use the scale to examine the relationships between vocational need strength, need satisfaction, locus of control and voluntary worker turnover. Voluntary turnover was defined as intent to leave one's job, the dependent variable. The Kostreva Vocational Attitude Scale (KVAS) was designed to measure intent-to-leave and Internal (I), Powerful Others (P) and Chance (C) locus of control, in Levenson's (1972) format, based on Rotter's (1966) theory and the 20 need factors of the Theory of Work Adjustment (Dawis et al. 1968). Hoyt reliability coefficients were .74 I, .80 P, .87 C and .95 for intent-to-leave. As predicted, KVAS sub-scales had loworder positive correlations to Levenson's I, P and C scales. The Minnesota Importance Questionnaire, assessing vocational need strength, the Minnesota Satisfaction Questionnaire, assessing vocational need satisfaction, and the KVAS, were administered to 101 randomly selected, employed persons, ages 20-40 (52 men, 49 women). Consistent with prior research, the intending-to-leave group had a significantly lower mean score on overall job satisfaction than the not-intending-to-leave group (p=.001). Step-wise regression analyses indicated significant correlations (p < .05) between: 1) low need satisfaction and intent-to-leave on 17 of 20 vocational need factors; 2) high need strength and intent-to-leave on the Advancement factor; and 3) high Chance locus of control and intent-toleave on the Social Status factor. One-way ANOVA showed a significantly higher mean score on overall Chance locus of control for the leaving group (p=.021), but no differences on Internal or Powerful Others. Post hoc discriminant analysis indicated younger age to be significantly related to intent-to-leave. Age and total job satisfaction correctly classified 75% of subjects to leaving and not-leaving groups (i.e., a younger individual who indicated low job satisfaction was likely to be in the leaving group). Gender and years of school were not related to leaving. Further study on actual turnover versus intent-to-leave was suggested, as was additional testing with the KVAS.