BACKGROUND: Healthy couple relationships are fundamental to a healthy society, whereas relationship breakdown and discord are linked to a wide range of negative health and wellbeing outcomes. Two types of relationship services (couple counselling and relationship education) have demonstrated efficacy in many controlled studies but evidence of the effectiveness of community-based relationship services has lagged behind. This study protocol describes an effectiveness evaluation of the two types of community-based relationship services. The aims of the Evaluation of Couple Counselling study are to: map the profiles of clients seeking agency-based couple counselling and relationship enhancement programs in terms of socio-demographic, relationship, health, and health service use indicators; to determine 3 and 12-month outcomes for relationship satisfaction, commitment, and depression; and determine relative contributions of client and therapy factors to outcomes. METHODS/DESIGN: A quasi-experimental pre-post-post evaluation design is used to assess outcomes for couples presenting for the two types of community-based relationship services. The longitudinal design involves a pre-treatment survey and two follow-up surveys at 3- and 12-months post-intervention. The study is set in eight Relationships Australia Victoria centres, across metropolitan, outer suburbs, and regional/rural sites. Relationships Australia, a non-government organisation, is the largest provider of couple counselling and relationship services in Australia. The key outcomes are couple satisfaction, relationship commitment, and depression measured by the CESD-10. Multi-level modelling will be used to account for the dyadic nature of couple data. DISCUSSION: The study protocol describes the first large scale investigation of the effectiveness of two types of relationship services to be conducted in Australia. Its significance lies in providing more detailed profiles of couples who seek relationship services, in evaluating both 3 and 12-month relationship and health outcomes, and in determining factors that best predict improvements. It builds on prior research by using a naturalistic sample, an effectiveness research design, a more robust measure of relationship satisfaction, robust health indicators, a 12-month follow-up period, and a more rigorous statistical procedure suitable for dyadic data. Findings will provide a more precise description of those seeking relationship services and factors associated with improved relationship and health outcomes.