This study was concerned with peer influence and social anxiety as predictors of self-medication. The participants consisted of 109 males (36.9%) and 186 females (63.1%). An adapted form of the Social Interaction Anxiety Scale (SIAS) and a self-developed instrument were used for the collection of data for the study. Cluster sampling and simple random sampling were the techniques and sampling methods adopted for this study. Three research hypotheses were posed and tested. The first stated that there would be no significant relationship among peer influence, social anxiety and self-medication. The second hypothesis stated that peer influence and social anxiety would have no significant joint and separate influence on self-medication. And the third stated that there would be no significant difference in self-medication based on gender. The result of the Correlation Matrix for Hypothesis one showed that there is a positive significant relationship among peer influence, social anxiety and self-medication at (r = 0.412, 0.191 & 0.363, df= 99, p<0.01). The result of the ANOVA analysis partially supported the second (F=30.315, df = 2, p <0.01) hypothesis. The result of the Independent T-test for hypothesis three revealed that there is no significant difference in self-medication based on gender with a t-observed value of 1.405. These findings reveal that there is a significant level of influence from peers and social anxiety on the activity of self-medication. More studies can be done in the area of comparing youths with other age groups in the prevalence of self-medication.