Over recent years there have been a number of initiatives for improving the quantity and quality of evaluation in public relations programmes. This has been partly because public relations practitioners wish to demonstrate their contribution, but also because, as expenditure on public relations grows (£3 billion worldwide by the end of 1999), clients and organizations are demanding to know that their money is being well-spent. To aid evaluation, a number of models have been developed. This paper critically analyses the most well-known and proposes a new context/objectives-driven model which combines aspects of previous models while refining and adding to them. Within this context, 25 UK award-winning public relations programmes were analysed in order to examine the types of evaluations they used. It was found that only 12 out of 103 evaluations used social science-based research methods for verifying their results. Fifty-one of the 103 evaluations measured programme impact with the other evaluations concentrated on programme outputs. The author found there were issues surrounding the framing of objectives which offered a partial explanation. A number of recommendations are made to programme specifiers and award competition organizers in order to help improve the quality of evaluations in the future.