For the first time since 1990 the Limfjord, which is the main Danish production area for mussels, was partly closed for a period of 14 wk during autumn 2006 because of presence of DSP toxins in blue mussels. The causative organism was Dinophysis acuminata. Only the okadaic acid (OA) and OA esters were observed. The maximum concentration level of approximately 600–700 μg OA equivalents kg⁻¹ was reached at approximately 5,000 cells l⁻¹ of D. acuminata. A further increase in the concentration of D. acuminata did not lead to any further increase in the concentration of OA in the blue mussels. The temporal development in the concentration of D. acuminata cells in the water was in good accordance with the development in the content of OA in blue mussels on a weekly basis. The maximum concentration of D. acuminata during the bloom was 14,000 cells l⁻¹ and the maximum concentration of total OA equivalents in the raw blue mussels was 706 μg/kg. Single cell toxicity measurements showed a mean content of 31 pg OA cell⁻¹ (min 19; max 72 pg OA cell⁻¹). The impact of the concentration procedure was investigated using a range of volumes filtered from 0,2 l–10 l filter⁻¹. Toxins were lost during the filtration process as a response to amount of water concentrated on each filter. The relationship between the volume filtered per filter and the content of OA observed in the cells fitted well the equation y = 32.0·e^−0.12·x, where y = OA cell⁻¹ (pg OA cell⁻¹) and x = volume of water filtered per filter (l). The OA content per cell measured when concentrating only 0.2 L per filter was a good measure of the theoretical content of toxins calculated in a “no handling” case. The influence of heat-treatment on the concentration of DSP toxins in mussels was investigated. Results of OA concentration in paired samples of raw and heat-treated blue mussels showed a 1.8 times higher concentration in heat-treated mussels than in raw blue mussels. In most official monitoring programs raw mussel tissue is used for the analysis of content of DSP toxins. However, most shellfish are eaten cooked or steamed so heat-treated tissue seems to be the most relevant matrix to analyze. This should be recognized in relation to future risk assessments to establish permissible levels of DSP toxins in shellfish and in relation to the control of the permissible levels.