Most depressed patients are seen and treated exclusively by primary care clinicians. However, primary care patients with depression are often not adequately treated. The aims of this pilot study were to measure the impact of a telephone disease management program on patient outcome and clinician adherence to practice guidelines, measure the relationship of clinician adherence to patient outcome, and explore the measurement of patient adherence to clinician recommendations and its impact on patient outcomes. Thirty-five primary care practices in the University of Pennsylvania Health System were randomized to telephone disease management (TDM) or "usual care" (UC). All patients received a baseline and a 16-week follow-up clinical evaluation performed over the telephone. Those from TDM practices also received follow-up contact at least every 3 weeks, with formal evaluations at weeks 6 and 12. These interval contacts were designed to facilitate patient and clinician adherence to a treatment algorithm based on the Agency for Health Research and Quality (AHRQ) practice guidelines. Depressive symptoms evaluated with the Community Epidemiologic Survey of Depression (CES-D) scale as well as guideline adherence were the primary outcome measures. Sixty-one patients were enrolled in this pilot project. The overall effect for CES-D scores over time was significant, (P <.001), indicating that those participating in the trial (both TDM and UC groups) showed significant improvement. The interaction between intervention condition and time was also significant (P <.05), indicating that TDM patients improved significantly more over time than did UC patients. A greater proportion of TDM patients had CES-D scores <16 by Week 16 (66.7 versus 33.3%; chi(2), P <.05). The improvement in depression outcome for the TDM group was related to its impact on improving clinician adherence to depression treatment algorithms. The TDM pilot did not show a statistically significant effect on improving patient adherence to clinician recommendations, however. This preliminary data suggests that TDM for depression improves both clinician guideline adherence and patient outcomes in the acute phase of depression. The effect on patient outcome is at least partially explained by the effect of TDM on clinician adherence to depression treatment algorithms.