During summer 2001, blue mussels Mytilus edulis with abnormal shell growth were collected near Kragerø, southern Norway. The mussels had green spots in their mantle tissues, mainly posteriorly and ventrally, and in the adductor muscle. Mussels from 4 sites had a prevalence of green spots varying from 2 to 71% that correlated well with shell deformities. Histological examination revealed the presence of round or ovoid algae, 0.9 to 1.5 × 1.2 to 2.4 µm, free within haemocytes and in the lesions, characterised by an inflammatory response and the presence of cellular debris. The alga contain a relatively large nucleus, 1 chloroplast and 1 mitochondrion. Size and morphology suggest that the alga might be a picoeucaryot green alga. Infection of mussel tissues appears to start in the posterior mantle edge, near the siphons, and spread anterior-ventrally in the mantle connective and storage tissues -occasionally spots were also found in the gonad follicles. Large infected areas were also observed in sinuses within the adductor muscle. Only mussels that were 3 yr old or more were infected. Deformations apparently resulted from years of continuous shell formation by a contracted, partly deformed mantle. Most deformed mussels had eroded shells, allowing some light penetration through the exposed, thin nacre. Young, thin-shelled mussels were not infected. The present work suggests that the alga has, at least partially, a parasitic relationship with the mussels, and is associated with pathological alterations in mussel tissues. MATERIALS AND METHODSField sampling. Mussels were sampled from 4 sites around Burøy on the island of Skåtøy, and Jomfruland, near Kragerø, southern Norway (58°50' N, 9°35' E) ( Fig. 1) in June 2002 and April 2003. Three samples (1a, 1b and 1c) were collected at Site 1, and 1 sample at each of Sites 2, 3 and 4. Each sample consisted of 150 randomly collected mussels, dredged or picked by hand. Sample 1c consisted of empty shells and shell fragments from approximately 300 mussels. Additional mussels were sampled for fixation of affected tissues. Geographical, hydrographical and biological conditions were noted at each sampling site.Gross morphology. During sampling, 20 mussels were steamed, aircooled, and opened to determine if green spots were apparent after processing. Samples 1a, 2, 3 and 4 were subsequently steamed and cooled separately. The length of each mussel shell was measured to the nearest 1 mm. Shell and tissue morphology was categorised as: shell morphology (normal/ deformed); infection site (adductor muscle/mantle edge/gonad area); and infection rate (no/low/moderate/high). Sample 1a had the highest prevalence of infected mussels, and was used for examination of the relationship between infection and shell length. Age of the mussels was estimated using Sample 1b, which was divided into mussel year-classes based on size, shell shape (fast versus slow growing specimens), presumed winter growth rings and erosion and fouling, partly according to Seed (1968). The length of each musse...