One of the fundamental limitations of atomic methods in biological ana
lysis is their inability to distinguish individual physico-chemical fo
rms of the metal. After a brief overview of 'hot' trace elements and a
tomic techniques used for total element analysis in bioanalytical work
, the importance and main challenges of speciation of toxic metals in
biological systems is addressed. The main analytical problems of speci
ation and present techniques/analytical strategies to tackle this prob
lem are highlighted. Recent work on metal speciation in our laboratory
is described in order to show that analytical difficulty is dependent
on the chemical nature of the sought species (i.e., moving from 'stab
le/kinetically inert' to 'unstable/fast reacting' species determinatio
ns). New analytical strategies for more stable species (e.g., methylme
rcury) by coupling a powerful separation technique with specific (atom
ic) detectors are described. The concept and analytical application of
non-chromatographic and vesicles-mediated HPLC-volatile species gener
ation-atomic detection to the speciation of toxic species of Hg, As or
Sn is discussed. It is emphasized that the complexity of toxic metal
speciation in biological matrices calls for a 'several-complementary'
analytical strategies approach. This concept of applying different-pri
nciple-based separation units (e.g., ultramicrofiltration, FI or HPLC
columns with different adequate packings) coupled with complementary d
etectors (usually atomic ones) for tackling complex problems is stress
ed. Comparative studies on the speciation of aluminium and silicon in
human serum carried out in the author's laboratory are used throughout
to illustrate this important point. Finally, some clinically relevant
conclusions derived from such trace metal speciation research are hig
hlighted.