Fine soil with a particle size less than 2 mm is generally used when determ
ining Bray-2 phosphorus (P). The average aggregate size varies widely in th
e fine soil of structural soils and may influence the reproducibility at Br
ay-2 extractions. Our objective was to study how this variation in fineness
is related to filtering time and jointly influence Bray-2 P values in thre
e different soil types and then to propose corrective measures to improve t
he reproducibility at Bray-2 extraction. In an Ultisol, there was no differ
ence in Bray-2 P values between soil sieved through a l-mm mesh compared to
soil sieved through a 0.2-mm mesh. Bray-2 P was drastically decreased in t
he fine (less than 0.2 mm) allophanic Andisols compared to soil passed thro
ugh a l-mm mesh. The difference disappeared gradually, when extending the e
xtraction time from the method specific for 40 or 60 seconds and was not ob
served at 3 min extraction time or longer. The finer structure delayed the
filtering time through filter papers which increased the resorption to allo
phanic soil constituents of the initially released P. Controlled filtering
time by using micropore filters revealed an initial dissolution of P. The s
orption rate became larger than the release rate after about 15 seconds, re
sulting in a net resorption in the allophanic Andisols. To the contrary, th
e Bray-2 P values increased for several minutes after addition of the extra
ctant to a nonallophanic Andisol. The P values increased faster with the fi
ner soil fraction compared to the coarser one. The variation in Bray-2 P va
lues caused by soil fineness can be reduced by controlling the filtering ti
me or prolonging the extraction time to 15 min. A prolonged extraction time
will facilitate the laboratory work by permitting analyzing in batches.