I also found that the MTT assay is not reliable. There is a Glo assay, I think from Sigma/aldrich that works nicely.

What do you precisely mean by "not so accurate": variations between replicates - or unexpected response of tetrazolium reduction by the drug? MTT is a (mainly mitochondrial) metabolic assay and may give you effects of drugs on metabolism not directly correlated with "viability". Have you done a cell count under the same conditions? Does that correlate with the MTT result?

If you want to use another  tetrazolium assay, try the water-solubles ones, such as WST-1. Since they are easier to handle, the assay may be less error-prone. Nevertheless, the above comments also apply as for MTT, except that WST-1 measures s cell surface NADH-oxidase. (see  Berridge MV, Herst PM, Tan AS. Tetrazolium dyes as tools in cell biology: new insights into their cellular reduction.

Biotechnology annual review 02/2005; 11:127-52. DOI:10.1016/S1387-2656(05)11004-7 Research Gate)

How much cells you initially use for testing, whether your drug interfering with MTT. Does you verify this assay with trypan/clonogenic assay..for any cell lines..? If no, then focus on these things..hope you will get result.

All the best.

 I am doing this assay in 96 well plate using 5000 cell per well. "Not so accurate" means the percentage cell viability is fluctuating in same concentration of compound up to 30%. Is this normal in this assay to have a big error bar?

Fluctuations of 30% are surely not acceptable and indicate that there may by an experimental error, either in your plating the cells (which type?) which will give you different cell numbers, and/ / or in performing the MTT assay, e.g. the complete solubilization step.

To continue Marci Moss's suggestion, CellTiter Glo from Promega is a very robust, high dynamic range, and very good reproducibility assay.  Things to note:

1.  You need a bioluminescence plate reader.

2.  It is more expensive than MTT or similar assays.  Promega is the cheapest vendor.

3.  But since it measures ATP, like MTT it is also a metabolic assay.

We have also used Presto Blue assay, with good effect, in the fluorescence mode. (It can also be done in Absorbance mode).  The big advantage is that it is a live-format assay (non-destructive), and one can continue growing cells in the presence of Presto Blue. Thus, you can follow cells in sample plate every day.  It is also more expensive, and since it is similar in concept to MTT, it may have similar variability.  But in fluorescence mode you have a stronger signal.

Good luck !

I am thankful to the all answers.

You can try also this one

https://fi.promega.com/resources/protocols/technical-bulletins/0/celltiter-glo-luminescent-cell-viability-assay-protocol/

There are various methods to evaluate cellular proliferation but at different endpoints.

[1] A simplest way is to directly count cell numbers at different days after plating.

[2] A colorimetric assay based on MTT, XTT, MTS, WST-1 or WST-1 measures the enzymatic activity of mitochondrial dehydrogenase (succinate-tetrazolium reductase) that converts these tetrazolium salts into purple colored insoluble formazan. A similar assay employing resazurin or AlamarBlue can be used both colorimetrically and fluorometrically. The assay of this type evaluates the mitochondrial metabolism.

[3] Tritiated thymidine uses radioisotope and evaluates the DNA synthesis. A similar assay employing other thymidine analogues BrdU, IdU, CldU or EdU does not need the use of radioisotope. There have been S phase markers available, such as Ki-67, PCNA, DNA topoisomerase II, and histone H3. Evaluation of E2F (required for a transition from G1 to S) or other proliferation-related markers may also be useful.

[4] A dye exclusion assay measures the membrane permeability and uses colorimetric (e.g., trypan blue, Erythrosin B) or fluorescent dyes (e.g., propidium iodide, ethydium bromide).

[5] A colony formation or clonogenic assay measures the ability of single cells to form colonies with 50 or more cells, usually in about two weeks after plating.