A Nursing Centre in Canada was initiated to demonstrate nursing practice in a primary health care context, unencumbered by conventional health care agency parameters. As one part of a multimethod evaluation, a 5 instrument client questionnaire was designed. The main instrument was developed by the researchers to measure the impact of the Centre and nurses' work from the perspective of clients. No established instrument could be found that was appropriate for the Centre's diverse client population and nature and variation in outcomes. To guide the development of the new instrument, minimally structured client interviews were conducted with 15 clients. To assess the validity and reliablity of other selected instruments for the population, clients were asked to comment on the appropriateness of items. Based on findings, the Nurses' Work and Client Outcomes Questionnaire (NWCOQ) was developed, the Client Satisfaction Questionnaire and demographic questions modified, and a Cantril Self-Anchoring ladder to assess health status was supported for use. An open section was added to invite further description of clients' experiences, in particular, the unexpected, contrary or outstanding. Further pilot work was done and the refined five instrument questionnaire then was used to survey all clients who used the Centre during a seven month period. Findings from the NWCOQ revealed the nature of the nurses' work and the degree change occured for clients on multiple health dimensions: physical, emotional, managing their situation, and group and community involvement. Clients were highly satisfied with the Centre and experienced many positive outcomes. The quantitative instruments and results were examined in relation to results from other data collection approaches and data sources. The NWCOQ has evidence of content validity and internal consistency reliability. The use of qualitative methods to develop and refine quantitative instruments for assessment of health outcomes for diverse client populations is highly recommended. This strategy is feasible even when outcome measurement timeframes are short.